fenvs-08-612253 December 1, 2020 Time: 20:27 # 1

ORIGINAL RESEARCH published: 07 December 2020 doi: 10.3389/fenvs.2020.612253

Environmental DNA (eDNA) Monitoring of Noble Crayfish Astacus astacus in Lentic Environments Offers Reliable Presence-Absence Surveillance – But Fails to Predict Population Density

Stein I. Johnsen1†, David A. Strand2*†, Johannes C. Rusch2,3 and Trude Vrålstad2

1 Norwegian Institute for Nature Research, Lillehammer, Norway, 2 Norwegian Veterinary Institute, Oslo, Norway, 3 Department of Biosciences, University of Oslo, Oslo, Norway

Noble crayfish is the most widespread native freshwater crayfish species in Europe. It is threatened in its entire distribution range and listed on the International Union for Edited by: Concervation Nature- and national red lists. Reliable monitoring data is a prerequisite for Ivana Maguire, University of Zagreb, Croatia implementing conservation measures, and population trends are traditionally obtained Reviewed by: from catch per unit effort (CPUE) data. Recently developed environmental DNA Michael Sweet, (eDNA) tools can potentially improve the effort. In the past decade, eDNA monitoring University of Derby, United Kingdom Chloe Victoria Robinson, has emerged as a promising tool for species surveillance, and some studies have University of Guelph, Canada established that eDNA methods yield adequate presence-absence data for crayfish. *Correspondence: There are also high expectations that eDNA concentrations in the water can predict David A. Strand biomass or relative density. However, eDNA studies for crayfish have not yet been [email protected] able to establish a convincing relationship between eDNA concentrations and crayfish †These authors have contributed equally to this work density. This study compared eDNA and CPUE data obtained the same day and with high sampling effort, and evaluated whether eDNA concentrations can predict Specialty section: relative density of crayfish. We also compared two analytical methods [Quantitative This article was submitted to Conservation and Restoration real-time PCR (qPCR) and digital droplet PCR (ddPCR)], and estimated the detection Ecology, probability for eDNA monitoring compared to trapping using occupancy modeling. In a section of the journal Frontiers in Environmental Science all lakes investigated, we detected eDNA from noble crayfish, even in lakes with very Received: 30 September 2020 low densities. The eDNA method is reliable for presence-absence monitoring of noble Accepted: 17 November 2020 crayfish, and the probability of detecting noble crayfish from eDNA samples increased Published: 07 December 2020 with increasing relative crayfish densities. However, the crayfish eDNA concentrations Citation: were consistently low and mostly below the limit of quantification, even in lakes with very Johnsen SI, Strand DA, Rusch JC and Vrålstad T (2020) Environmental high crayfish densities. The hypothesis that eDNA concentrations can predict relative DNA (eDNA) Monitoring of Noble crayfish density was consequently not supported. Our study underlines the importance Crayfish Astacus astacus in Lentic Environments Offers Reliable of intensified sampling effort for successful detection of very low-density populations, Presence-Absence Surveillance – But and for substantiating presumed absence, inferred from negative results. Surprisingly, Fails to Predict Population Density. we found a higher likelihood of eDNA detection using qPCR compared to ddPCR. Front. Environ. Sci. 8:612253. doi: 10.3389/fenvs.2020.612253 We conclude that eDNA monitoring cannot substitute CPUE data, but is a reliable

Frontiers in Environmental Science| www.frontiersin.org 1 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 2

Johnsen et al. eDNA Monitoring of Noble Crayfish

supplement for rapid presence-absence overviews. Combined with eDNA analyses of alien crayfish species and diseases such as crayfish plague, this is a cost-efficient supplement offering a more holistic monitoring approach for aquatic environments and native crayfish conservation.

Keywords: noble crayfish Astacus astacus, eDNA, occupancy modeling, qPCR, ddPCR, CPUE, detection frequency, relative density

INTRODUCTION a filter or pelleted from a water sample, from which DNA can be extracted and analyzed. Thus, from a water sample, it is Freshwater crayfish are regarded as keystone species and shape possible to detect specific species or even whole communities the littoral zone in both lotic and lentic environments (Creed, (Deiner et al., 2017; Bylemans et al., 2019; McElroy et al., 2020). 1994; Momot, 1995). Their presence in aquatic environments, Quantitative real-time PCR (qPCR) or digital droplet PCR influencing sediment dynamics and benefiting other animals, (ddPCR) are commonly used for species-specific detection has also led freshwater crayfish to be characterized as ecosystem (Rusch et al., 2018; Capo et al., 2019; Mauvisseau et al., 2019a; engineers and umbrella species (Usio and Townsend, 2001; Strand et al., 2019), and for relative or absolute quantification of Reynolds et al., 2013). Furthermore, they are regarded as target DNA, respectively (Demeke and Dobnik, 2018; Quan et al., indicator species for water quality (Sylvestre et al., 2002). In 2018), while high-throughput sequencing and metagenomics addition, some species of freshwater crayfish are harvested and is used to study whole communities (Thomsen et al., 2012; regarded as delicacies, obtaining high prices in the market Hänfling et al., 2016; McElroy et al., 2020). While qPCR is (Ackefors, 1998). One of these prized species is the noble crayfish, currently the most common platform to analyze eDNA samples Astacus astacus, which is indigenous to Europe and the only using species-specific assays, recent studies suggest that the indigenous species of freshwater crayfish in Norway (Souty- detection rate of eDNA in environmental samples is higher when Grosset et al., 2006; Kouba et al., 2014). There are currently about using ddPCR compared to qPCR technology (Doi et al., 2015a; 470 registered Norwegian populations of noble crayfish (Johnsen Mauvisseau et al., 2019a; Wood et al., 2019; Brys et al., 2020). and Vrålstad, 2017). Along with populations of other native With ddPCR there is no need for standard curves and ddPCR freshwater crayfish species indigenous to Europe, the number allows for absolute quantification, even at low levels of DNA of noble crayfish populations has declined dramatically in the copies. Additionally, ddPCR appears to be more robust against last decades. This is mostly due to introduced North-American PCR inhibition (Doi et al., 2015a; McKee et al., 2015; Mauvisseau crayfish species that carry and transmit the crayfish plague et al., 2019a; Brys et al., 2020). pathogen Aphanomyces astaci, but also due to anthropogenic Environmental DNA methods allow for rapid detection of influences such as pollution and habitat loss (Holdich et al., targeted species and are considered a promising monitoring tool 2009; Kouba et al., 2014). Hence, the noble crayfish is both on for aquatic species surveillance and inventories (Lodge et al., the international (Edsman et al., 2010) and the national red 2012; Thomsen and Willerslev, 2015), including monitoring and list (Henriksen and Hilmo, 2015). The red list status and its discovery of endangered and invasive species (Dejean et al., 2012; importance in freshwater ecosystems has led to the development Laramie et al., 2015; Strand et al., 2019). Additionally, eDNA of surveillance programs aiming to monitor distribution and methodology is non-invasive compared to more traditional relative density of noble crayfish (Johnsen et al., 2019). In methods where the species itself is caught (Thomsen and Norway, as in other countries, estimates of relative density Willerslev, 2015). An increasing number of studies show that are obtained by trapping crayfish with baited traps (Johnsen eDNA monitoring is suitable for acquiring presence/absence et al., 2014). This is relatively time consuming, and in order to information of targeted species (Hempel et al., 2020; Mason increase the number of monitored populations, environmental et al., 2020; Villacorta-Rath et al., 2020), although “not-detected” DNA (eDNA) methodology has recently been included in the data cannot be taken as absolute proof of absence (Rusch et al., Norwegian surveillance programs for both crayfish plague and 2020). For marine and freshwater fish species, it has also been freshwater crayfish (Johnsen et al., 2019; Strand et al., 2020). The suggested that eDNA concentrations can be used to estimate methods are also used in crayfish monitoring studies in Europe population density or biomass of a species (Takahara et al., (Robinson et al., 2018; Mauvisseau et al., 2019b; Rusch et al., 2020; 2012; Thomsen et al., 2012; Lacoursière-Roussel et al., 2016). For Troth et al., 2020). freshwater crayfish, several studies demonstrate the use of eDNA During the past decade, eDNA methods have been to determine the presence or absence of species (Tréguier et al., increasingly used as monitoring tool for freshwater organisms 2014; Agersnap et al., 2017; Harper et al., 2018; Strand et al., 2019; (Leese et al., 2016; Bylemans et al., 2019; Strand et al., 2019; Rusch et al., 2020), which is very useful for verification of species Goutte et al., 2020). These methods utilize DNA traces in the presence and species distribution. A few studies have investigated environment originating from single-celled microorganisms or the relationship between population density of targeted crayfish cells shed from complex organisms in the form of propagules, species and eDNA concentration, but so far no or only weak mucus, abraded epithelial cells and body fluids (Thomsen and correlations have been reported (Dougherty et al., 2016; Cai et al., Willerslev, 2015). These sources of eDNA are easily caught on 2017; Larson et al., 2017; Rice et al., 2018; Troth et al., 2020). More

Frontiers in Environmental Science| www.frontiersin.org 2 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 3

Johnsen et al. eDNA Monitoring of Noble Crayfish

studies are thus needed to evaluate whether eDNA concentrations setting the traps, but still on the same day. The samples were can somehow reflect the relative population density of freshwater filtered directly from the boat using a battery driven peristaltic crayfish species. pump (ES portable sampler, Masterflex, Cole-parmer, Vernon Hills, United States) with tygon tubing (Masterflex) and a 47 mm Goal of Study inline filter-holder (Millipore, Billerica, United States) (Strand We aimed to make a direct comparison between the et al., 2019). Glass-fiber filters (AP2504700, Millipore) with an traditional monitoring of noble crayfish using baited traps, effective pore size of 2 µm were used, as in previous studies and targeted eDNA monitoring by means of species-specific (Agersnap et al., 2017; Strand et al., 2019; Rusch et al., 2020). qPCR and ddPCR. This will help assess if eDNA yields valid The inlet of the tube was attached to a small plastic box weighed presence/absence data for noble crayfish. We further use down with lead in order to collect water ∼5 cm above the bottom, occupancy modeling to estimate the probability to detect eDNA between 1 and 3 meters depth depending on the shoreline and of noble crayfish at various population densities, ranging from lake. After filtration, each filter was transferred to a 15 ml falcon very low to high density populations. We also explore whether tube and stored on ice until return to the laboratory where the eDNA concentrations in the water correlates with observed filter samples were frozen awaiting DNA extraction. relative density of noble crayfish. Finally, we compared the qPCR and ddPCR results and efficiency for eDNA detection Extraction of gDNA and eDNA of noble crayfish. In order to obtain reference DNA for positive controls and standard dilution series, genomic DNA (gDNA) was extracted from noble crayfish tissue (abdominal muscle) MATERIALS AND METHODS using QIAmp R DNA mini kit with the QIAcube automated DNA extractor (Qiagen) following the manufacturer’s protocol. Study Sites The DNA concentration was measured using the Qubit 1x In total eight lakes in the south-eastern part of Norway were dsDNA HS Assay Kit and Qubit 4 Fluorometer (Invitrogen, included in this study (Figure 1A and Table 1). These study life technologies) according to the manufacturer’s protocol. sites were selected according to an expected range of crayfish Environmental DNA was extracted from the glass fiber filters abundance, based on results from other projects such as the using an extraction protocol as described in Strand et al.(2019). National surveillance program for noble crayfish (Johnsen et al., In short, eDNA samples were lysed in cetyltrimethylammonium 2019). Here, we have available data from several years of trap buffer (CTAB) and proteinase K at 65◦C for 60 min, catches, including estimates of relative density from catch per cleaned and separated using chloroform, precipitated using unit effort data (CPUE). To avoid larger molting periods, which isopropanol and re-suspended in TE-buffer (pH 8). Due would influence both the catchability of crayfish (Westman and to the large volume of eluate (4 ml) from each sample, Pursiainen, 1982) and possibly the eDNA concentrations in the the eluates were divided into two subsamples to bypass water (Laurendz, 2017), six of the locations were sampled after the volume restrictions caused by centrifuge size. These mid-August. Due to the high fishing pressure in Lake Einafjorden subsamples were then merged after re-suspension in TE- and Steinsfjorden, we surveyed these two localities from the 7 to buffer. During the extraction of DNA, an environmental blank 9th of August 2016. control and an extraction blank control were incorporated in order to measure potential contamination during the DNA Crayfish Trapping extraction step. In each lake, 50 funnel LiNi traps with two entrances and 14 mm mesh size were set along the shoreline in the depth interval 0.5– qPCR and ddPCR Protocols 5 m (Table 1 and Figure 1B), in accordance with the methods gDNA and eDNA extracts were analyzed with a species-specific used in the national monitoring program for noble crayfish assay targeting the cytochrome c oxidase subunit I (COI) of (Johnsen et al., 2019). The traps were baited with raw chicken, noble crayfish published in Rusch et al.(2020): forward primer set in the evening and emptied the next morning. The relative 50-CCC CTT TRG CAT CAG CTA TTG-30, reverse primer density or CPUE is given as the number of crayfish per trapnight. 50-CGA AGA TAC ACC TGC CAA GTG T-30 and probe The number of crayfish caught per trapnight is regarded as a FAM-50 CTC ATG CAG GCG CAT-MGBFNQ. This assay was reliable measure of crayfish density with sufficient effort, and even specifically designed and optimized for both the qPCR and at efforts as low as 15 trapnights (Zimmerman and Palo, 2011; ddPCR platforms. A total of six technical replicates (three Johnsen et al., 2014). undiluted and three 5x diluted) were analyzed from each sample on both platforms (qPCR and ddPCR). The qPCR analysis was Water Samples performed on a Bio-Rad CFX96 Touch (Bio-Rad, Hercules, A total of 8–12 water samples were collected for eDNA analysis CA, United States) in a 25 µL reaction volume. Each reaction from each lake, along the same transect in which the traps were consisted of 12.5 µL TaqMan Environmental Master Mix 2.0 placed (Figure 1B). We filtered up to 5 L per sample, but in (Thermo Fisher Scientific, Waltham, United States), 500 nM some cases less if the filter clogged. In order to avoid any possible of each primer, 250 nM probe, nuclease free water and 5 µl contamination or influence of eDNA results from the trapping DNA sample. The thermocycling protocol consisted of an initial activity, the water samples were collected and filtered prior to warming at 95◦C for 10 min followed by 50 cycles of 95◦C for

Frontiers in Environmental Science| www.frontiersin.org 3 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 4

Johnsen et al. eDNA Monitoring of Noble Crayfish

FIGURE 1 | (A) Location of the monitored lakes in Norway. Black dots indicate the lakes and numbers identify the lakes, referring to Table 1. (B) Illustration of sampling transect of crayfish trapping and eDNA sampling at location 8. Here, 50 traps were set along the shoreline in the gray shaded area, and 8–10 filtrated water samples (black dots) were collected from each lake.

TABLE 1 | Lake and lake size with sample dates and effort of eDNA samples and traps.

Lake (number) Lake size (ha) Sample date # eDNA samples Trap effort (trapnights)

Baereia (3) 138 01.09.2015 12 50 Baereia (3) 138 23.08.2017 10 50 N. Billingen (2) 176 02.09.2015 9 50 N. Billingen (2) 176 24.08.2017 10 50 S. Billingen (1) 136 02.09.2015 10 50 S. Billingen (1) 136 25.08.2017 10 50 Skårillen (4) 46 03.09.2015 10 50 Skårillen (4) 46 22.08.2017 8 50 Harestuvatnet (5) 206 10.09.2015 9 50 Gjerdingen (6) 300 09.09.2015 9 50 Steinsfjorden (7) 1384 09.08.2016 8 48 Einafjorden (8) 1382 07.08.2016 10 50

Lake number (within parenthesis is related to the number seen in Figure 1).

15 s and 60◦C for 60 s. Four known concentrations of 10-fold The ddPCR analysis was performed on a QX200 AutoDG diluted genomic DNA (noble crayfish) were run in duplicates as Droplet Digital PCR System (Bio-Rad) in a 22 µl reaction a positive control and standard with known DNA copy numbers volume. Each reaction consisted of 11 µl ddPCR supermix for (based on absolute quantification of ddPCR) in order to calculate probes (no dUTP, Bio-Rad), 900 nM of each primer, 300 nM the DNA copies in each reaction (relative quantification) using of probe, 0.6 µl bovine serum albumin (BSA), nuclease free the manufactures software (CFX Manager v. 3.1.1517.0823). water and 5 µl DNA. Droplets were generated using AutoDG

Frontiers in Environmental Science| www.frontiersin.org 4 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 5

Johnsen et al. eDNA Monitoring of Noble Crayfish

instrument (Bio-Rad), where an emulsion is created with 20 µl for each sample, and all analyzed blank controls were negative of the 22 µl reaction volume, resulting in a 10% loss of DNA (N = 81), we scored a sample positive when three or more of template and supermix. After droplet generation, the plates the six technical ddPCR replicates of the sample had one or were transferred to a TM100 thermocycler (Bio-Rad) with the more positive droplets. Reactions with total droplet count below following cycling conditions: an initial warming at 95◦C for 8,000 were excluded. 10 min followed by 45 cycles of 94◦C for 30 s and 60◦C for 60 s We estimated the DNA copy numbers per liter water from and a final step at 98◦C for 10 min to inactivate the enzyme. each sample for both qPCR and ddPCR according to Agersnap ◦ ∗ Ramp rate was set to 2 C/s. The plate was thereafter transferred to et al.(2017) using the equation: C L = [Cr (Ve/Vr)]/Vw. Here the QX200 droplet reader (Bio-Rad) for final analysis. A known CL = copies of eDNA per liter lake water, Cr = copies of eDNA concentration of genomic DNA (noble crayfish) was run in in reaction volume, Ve = total elution volume after extraction, duplicate as a positive control. The DNA copies in each reaction Vr = volume of eluded extract used in the qPCR/ddPCR reaction, were calculated (absolute quantification) using the manufactures Vw = volume of filtered lake water. software (Quantasoft v.1.7.4.0917). The environmental blank control and extraction black control Statistical Analysis from each batch of extracted DNA were included in both the The statistical analyses were performed using RStudio qPCR and ddPCR analysis. Negative blank controls (MilliQ (v.1.3.1073) and R (v.4.0.2). We compared the qPCR and ddPCR water) were also included on each plate analyzed. reaction estimates of DNA copies from gDNA using a spearman correlation test, and used a generalized linear model (GLM) Limits of Detection and Quantification with the binominal family with logit link logistic to compare For the qPCR approach, the limit of detection (LOD) had been detection (TRUE/FALSE) against method (qPCR/ddPCR) and established by Rusch et al.(2020) as five copies/qPCR reaction, dilution series. We compared the detection frequency of qPCR and followed the criteria that LOD is the lowest concentration and ddPCR of DNA from the field samples using a GLM with that yields a detection probability of 95%, ensuring <5% false the binominal family with logit link logistic to compare detection negatives (Berdal et al., 2008; Vrålstad et al., 2009). Likewise, (TRUE/FALSE) against method (qPCR/ddPCR) and CPUE. the limit of quantification (LOQ) for the qPCR assay has been We used the R package eDNA occupancy to run multiscale established as 10 DNA copies/qPCR reaction (Rusch et al., 2020), occupancy modeling in order to estimate the detection using the same acceptance level as set for qPCR quantification of probability of crayfish using the targeted eDNA approach the crayfish plague pathogen A. astaci (Vrålstad et al., 2009), with (Dorazio and Erickson, 2018). The multiscale occupancy model observed standard deviation <0.5 for the Ct-values. Following estimates (1) the probability of species occurrence at the location previous recommendations, a cut-off was set at Ct 41 for the (psi or ψ), (2) the conditional probability of occurrence in a qPCR assay (Agersnap et al., 2017; Strand et al., 2019; Rusch water sample given that the species is present at that location et al., 2020), implying that amplification at or above this value was (theta or θ), and (3) the conditional probability of detection in a not considered a positive detection. For each eDNA sample, the PCR reaction given that the species is present in the sample (ρ) following criteria for a positive sample were used: if three or more (Dorazio and Erickson, 2018). The models were run separately of the six technical qPCR replicates for a sample were positive for the qPCR and ddPCR data sets. For samples with less than below Ct 41, the sample was considered positive. three positive technical replicates all replicates were scored as The same standard dilutions used to estimate the LOQ and negative, as described above. We first tested the occupancy LOD for the qPCR assay in Rusch et al.(2020) were analyzed models with assumed constant parameters. Then we included the with the ddPCR assay in order to estimate the absolute limit factor CPUE into the occupancy modeling in order to investigate of quantification (aLOQ) and LOD. The aLOQ for ddPCR can how relative crayfish densities influence the detection probability be estimated as the lowest copy number of a target within in a water sample (θ). We assumed constant parameters for the dynamic range with a relative standard deviation (RSD) eDNA occurrence in lakes (psi) and in PCR replicate (p) while of the measured copy number ≤25% (Dobnik et al., 2015). the conditional probability of eDNA occurrence in water samples The theoretical LOD, i.e., the lowest concentration that yields a (theta) was assumed to be a function of CPUE. The models detection probability of 95%, is estimated to be 0.29 copies/µl were run for a total of 11,000 iterations. We used the equation for ddPCR in the case 17,000 accepted droplets and corresponds 1 − (1 − θ)n ≥ 0.95 to calculate the number of water samples to ∼5.8 copies in a 20 µl reaction (Pecoraro et al., 2019). The to achieve a 95% detection probability, where θ is the estimated standard used had a starter concentration of 50 ng/µl of noble probability to detect crayfish eDNA in a water sample. We also crayfish gDNA, and was four-fold diluted to create a dilution compared (spearman correlation) the mean DNA copies from − − series ranging from 12.5 ng/µl (4 1) to 0.00000075 ng/µl (4 13). each field sample from both qPCR and ddPCR with CPUE. As ddPCR offers absolute quantification, the ddPCR results from the 3rd fourfold-dilution of gDNA were used as a baseline for the qPCR standard in order to estimate the relative DNA copies in RESULTS each qPCR reaction. There is no consensus on the number of positive droplets that are required to score a ddPCR replicate Comparisons of qPCR and ddPCR Assay as positive, but the threshold is commonly set at 2, 3, or 5 Both the qPCR and ddPCR amplified the gDNA standard and droplets (Dobnik et al., 2015). Since we analyzed six replicates there is a significant correlation (R2 = 0.9, p < 2.2e-16) between

Frontiers in Environmental Science| www.frontiersin.org 5 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 6

Johnsen et al. eDNA Monitoring of Noble Crayfish

FIGURE 2 | (A) Correlation plot of the estimated DNA copies per reaction of qPCR and ddPCR analysis of gDNA from crayfish tissue using spearman correlation (R2 = 0.9, p < 2.2e-16). (B) Logistic regression curve showing probability of gDNA detection against the dilution series number using ddPCR. (C) Logistic regression curve showing probability of detection against the gDNA dilution series number using qPCR.

TABLE 2 | The qPCR and ddPCR results for the analysis of the standard dilution used to estimate LOQ and LOD for qPCR (Rusch et al., 2020) and ddPCR (this study) using the same assay.

qPCR ddPCR

Standard DNA N Cq Cq DNA copies SD** Detection RSD*** N Pos. DNA SD** Detection RSD*** (ng/µl) Std Drops* copies

Undiluted 5.00E + 01 S4∧1 1.25E + 01 6 18.00 0.07 3046534.65 139026.71 100% 5% 6 17101.3 >200000 NA NA NA S4∧2 3.13E + 00 8 19.95 0.06 798995.10 34118.09 100% 4% 8 18460.0 >200000 NA NA NA S4∧3 7.81E-01 8 22.01 0.04 194510.60 5600.27 100% 3% 8 18258.1 167000.00 7618.77 100% 5% S4∧4 1.95E-01 8 24.08 0.07 47080.87 2210.02 100% 5% 8 15028.3 40677.50 736.98 100% 5% S4∧5 4.88E-02 8 26.11 0.05 11668.85 377.37 100% 3% 8 6326.5 10017.50 226.89 100% 2% S4∧6 1.22E-02 8 28.09 0.07 3006.44 138.10 100% 5% 8 1890.1 2548.00 50.55 100% 2% S4∧7 3.05E-03 8 30.13 0.08 740.64 38.82 100% 5% 8 486.6 650.00 33.87 100% 5% S4∧8 7.63E-04 8 32.14 0.17 186.55 21.18 100% 11% 8 123.1 171.50 13.26 100% 8% S4∧9 1.91E-04 20 34.20 0.32 46.23 9.18 100% 20% 20 33.9 43.25 3.01 100% 7% S4∧10 4.77E-05 20 36.22 0.39 11.62 2.62 100% 23% 20 6.7 9.54 3.43 100% 36% S4∧11 1.19E-05 20 38.83 1.01 2.35 1.55 100% 66% 20 1.7 1.98 1.42 85% 72% S4∧12 2.98E-06 20 39.71 0.56 0.39 0.60 35% 156% 20 1.0 0.69 0.80 45% 115% S4∧13 7.45E-07 20 39.96 0.03 0.17 0.36 20% 205% 12 1.0 0.25 0.59 17% 234%

*Average number of positive droplets, negative droplets excluded. **Standard deviation. ***Relative standard deviation.

the estimated DNA copies from the two methods (Figure 2A). 42.6% of the 690 qPCR reactions had a positive amplification There was no significant difference (z = 0.492, p = 0.623, below the set cut-off (Cq 41). Both for the qPCR and ddPCR logistic regression) in detection rate between the two methods results, most of the reactions were below the LOQ (Figure 3) on the dilution series (Figures 2B,C). The aLOQ for the ddPCR indicating low concentration of noble crayfish eDNA. None of was estimated to be ∼ 40 copies, the last standard dilution the environmental blank controls or extraction blank controls where the RSD was ≤25% (Table 2). In the initial ddPCR run displayed any amplification or produced positive droplets. several of the positive field samples showed signs of inhibition Neither did the negative blank controls display amplification or and lower amplitude of the positive droplets compared to the produce positive droplets. positive control. Adding BSA to the reaction or diluting the sample (1:5) appeared to improve this issue by increasing the amplitude enough to separate negative from positive droplets. Relative Density of Crayfish Compared A total of 690 reactions were run both for qPCR and for ddPCR. to eDNA Presence/Absence Data Here, 33 ddPCR reactions were excluded due to total droplet The CPUE of noble crayfish in the different localities ranged count below 8,000. Of the remaining 657 ddPCR reactions, in mean values between 0.08 and 17.52 (Figure 4 and 28.5% of the reactions had one or more positive droplets, while Supplementary Table 1). According to the classification system

Frontiers in Environmental Science| www.frontiersin.org 6 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 7

Johnsen et al. eDNA Monitoring of Noble Crayfish

FIGURE 3 | A scatterplot of the estimated DNA copies per reaction of qPCR and ddPCR analysis of eDNA. The stippled line indicates a 1:1 ratio, while the horizontal line and vertical line indicates the LOQ for ddPCR and qPCR, respectively.

of the national surveillance program of noble crayfish in and the detection frequency increased (z = 4.55, p = 5.46e-06, Norway (Johnsen et al., 2019), the crayfish populations cover logistic regression) for both qPCR and ddPCR with increased all categories, ranging from “very low density” populations CPUE (Figure 5B). to “very high density” populations (Figure 4). We detected eDNA from noble crayfish in all lakes, but in 2015 we failed Occupancy Modeling to detect noble crayfish eDNA in one of the sampled lakes The occupancy models with constant parameters (Table 3) (Lake Skårillen). In this case, there was very heavy rainfall also showed that it is more likely to detect eDNA of crayfish before and during the sampling. Noble crayfish eDNA was using qPCR compared to ddPCR. Including the CPUE as a detected in this lake in 2017 (Figure 4). Even in lakes with factor in the occupancy models gives an estimate of detection very low crayfish densities (Figures 4, 5), noble crayfish eDNA likelihood in water samples at different relative densities of had been reliably detected with a qPCR detection frequency crayfish, and the likelihood to detect eDNA of noble crayfish ranging from 0.3 to 0.7. The detection frequency of noble in water samples increases with relative densities, both for crayfish eDNA was significantly higher (z = 4.27, p = 1.93e-05, qPCR and for ddPCR (Figure 6). At all but the highest relative logistic regression) for qPCR compared to ddPCR (Figure 5A), density (CPUE ∼17.52) estimated occurrence of eDNA in a

Frontiers in Environmental Science| www.frontiersin.org 7 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 8

Johnsen et al. eDNA Monitoring of Noble Crayfish

FIGURE 4 | Boxplot showing the number of noble crayfish per trapnight (CPUE) in the eight localities. The plots are based on the number of crayfish caught in the individual traps within the lake. The box contains 50% of the individual trap values, the median and the average values are shown as black and stippled red lines, respectively. The whiskers indicate the 5 and 95 percentiles and the dots (•) show values outside this interval. The pie charts depict the qPCR detection frequency (black part being positive) from the water samples. ddPCR detection frequency is not included. Exact values are given in Supplementary Table 1.

FIGURE 5 | (A) A comparison of the observed detection frequency (proportion of positive eDNA samples in a lake) between qPCR and ddPCR. The dotted line shows the 1:1 ratio. (B) A comparison of the observed detection frequency between CPUE against ddPCR and qPCR.

water sample is lower using ddPCR analysis compared to qPCR samples are sufficient to detect eDNA of noble crayfish with (Figure 6). For qPCR, the probability of eDNA occurrence 95% likelihood using qPCR, while 11 to 14 water samples is relatively high (around 0.5) even at the CPUE estimates are needed to detect noble crayfish with 95% likelihood using below 1 (Figure 6). At very low to low densities five water ddPCR (Figure 7).

Frontiers in Environmental Science| www.frontiersin.org 8 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 9

Johnsen et al. eDNA Monitoring of Noble Crayfish

TABLE 3 | The estimated likelihood of occurrence at the location (ψ), in the water sample (θ) or PCR replicate (p) for qPCR and ddPCR using fixed parameters.

psi (ψ ) theta (θ ) P

qPCR 0.876 (0.645–0.981) 0.542 (0.447–0.634) 0.706 (0.655–0.753) ddPCR 0.814 (0.563–0.965) 0.318 (0.225–0.419) 0.737 (0.671–0.799)

FIGURE 7 | Estimated number of water samples needed to achieve 95% detection likelihood at different relative densities of crayfish (CPUE) for qPCR and ddPCR.

where eDNA copy numbers calculated below LOQ are correlated with density (or biomass) estimates. We did a correlation test (spearman) in order to compare DNA concentration in water FIGURE 6 | The estimated probability of eDNA occurrence in water samples (DNA copies per liter) to relative density of crayfish (CUPE), (θ) using occupancy models where the probability of eDNA occurrence in lakes and the test showed a very weak positive correlation between was assumed to be constant, the conditional probability of eDNA occurrence eDNA concentrations and CPUE data, both for qPCR and ddPCR in water samples was assumed to be a function of CPUE and the conditional results (Figure 8). This result is highly uncertain and interpreted probability of eDNA detection in PCR replicate was assumed to be constant. with great caution.

Relative Density of Crayfish Compared DISCUSSION to eDNA Copy Numbers The copy numbers estimated from the qPCR and ddPCR Populations of native European noble crayfish are currently being reactions in our study were very low and most of the numbers lost at an alarming rate, largely because of North-American were below the LOQ (Figure 3). The estimated DNA copy invasive crayfish that carry and transmit the crayfish plague numbers that are below LOQ are unreliable due to high variation pathogen (Holdich et al., 2009; Kouba et al., 2014). There is when the copy numbers in a sample are low, and thus not suitable therefore an urgent need for better, powerful and dedicated for statistical analysis. However, it is not uncommon to see studies conservation and management strategies. This requires solutions

FIGURE 8 | (A) A scatterplot with correlation (Spearman) of the mean DNA copies per liter for each sample from ddPCR plotted against CPUE of crayfish. The horizontal line indicates the LOQ of ddPCR adapted to fit the copies per liter values. (B) A scatter plot with correlation (Spearman) of the mean DNA copies per liter for each sample from qPCR plotted against CPUE of crayfish. The horizontal line indicates the LOQ of qPCR adapted to fit the copies per liter values.

Frontiers in Environmental Science| www.frontiersin.org 9 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 10

Johnsen et al. eDNA Monitoring of Noble Crayfish

derived from understanding, knowledge and measures related from crayfish tissue, we had lower detection sensitivity for to biology, socioeconomics and legal frameworks. Here, efficient ddPCR compared to qPCR for the eDNA samples. This contrasts monitoring strategies play a key role. In the past decade, other studies where ddPCR appears to be more sensitive than eDNA monitoring has emerged as a promising, cost-effective qPCR for eDNA detection (e.g., Doi et al., 2015b; Hamaguchi tool for monitoring of endangered and invasive species (Jerde et al., 2018; Mauvisseau et al., 2019a; Wood et al., 2019; Brys et al., 2011; Lodge et al., 2012) and disease pathogens (Strand et al., 2020). In some eDNA samples we observed a reduced et al., 2011, 2014; Gomes et al., 2017) with the potential to amplitude for the ddPCR indicating some inhibition. Although simplify and streamline species surveillance (Kelly et al., 2014; ddPCR is robust against inhibition, high concentrations of PCR Leese et al., 2016). High expectations have been connected inhibitors like humic acids may also inhibit ddPCR. Several qPCR to eDNA data as predictors for species distribution, relative mastermixes are robust against inhibition, including Taqman density and biomass estimates (Bohmann et al., 2014; Kelly Environmental Mastermix 2.0 (Strand et al., 2011; Uchii et al., et al., 2014; Doi et al., 2017). However, with the growing body 2019). Thus, the type of qPCR mastermix and its ability to deal of reports and research, the first optimistic prospects regarding with inhibition will highly influence the results in a comparison the applicability of eDNA prediction have encountered reality, to ddPCR. Strand et al.(2011) found that Taqman Environmental turning out to vary for different groups of organisms, habitats and Mastermix removed close to 100% of the observed inhibition choice of methods. Several proof-of-concept studies have shown in DNA extracts for different natural water samples including that eDNA monitoring yields adequate presence-absence and water with high content of humic substances. For ddPCR, we distribution data for crayfish, with a high potential for crayfish experienced that the relevant ddPCR chemistry (Bio-Rad) had no surveillance. But, in contrast to some studies focusing on fish specific mastermix developed to deal with inhibitory substances (Lacoursière-Roussel et al., 2016; Doi et al., 2017), there are so far in the ddPCR other than the inherent property of ddPCR few if any studies on crayfish that have established a convincing (partitioning the samples into droplets), making it less sensitive to quantitative relationship between eDNA concentrations in the inhibition. Since noble crayfish habitats in the Northern countries water and crayfish biomass or density. We have systematically commonly are within or in close vicinity to boreal coniferous compared traditional trapping methods and eDNA monitoring forests, the water is commonly rich in humic substances. We with a thorough sampling effort. In concordance with other believe this could have negatively influenced the detectability for studies (Dougherty et al., 2016; Cai et al., 2017; Larson et al., ddPCR in this study. Though we did not include an internal 2017; Strand et al., 2019; Troth et al., 2020), our results support PCR control (IPC), it would be beneficial to use in this type of the assumption that eDNA yields valid presence-absence data, comparison (Goldberg et al., 2016). even at very low population densities. However, lakes with low-density crayfish populations require higher sampling effort than commonly used in comparable studies and monitoring eDNA Detection Probability Compared to programs. We found that the probability of detecting noble Crayfish Population Density crayfish from eDNA samples increased with increasing relative While we reliably detected noble crayfish eDNA in all localities crayfish densities, corroborating similar studies (Dougherty et al., with very low crayfish densities, the detection frequency was often 2016; Larson et al., 2017). However, no significant correlation very low. To achieve a 95% detection likelihood, low-density between noble crayfish eDNA concentrations in the water and lakes required an estimated five filter samples, corresponding to relative crayfish density could be established. Thus, our eDNA ∼25 L of water. Thus, the anecdotal “cup of water” (Lodge et al., data does not support the hypothesis that eDNA concentrations 2012) or rapid, low-volume sample effort seems insufficient for can predict relative crayfish density. monitoring low-density crayfish populations. Many other studies have also reported successful eDNA detection of freshwater crayfish at very low densities (Dougherty et al., 2016; Larson et al., Efficiency for eDNA Detection of Noble 2017; Strand et al., 2019), but often with low detection frequency. Crayfish – qPCR Versus ddPCR Poor detection efficiency can in some studies also be associated It is commonly reported that ddPCR overcomes the challenges with very small water volumes per sample (Tréguier et al., 2014; connected to inhibitory substances to a higher extent than qPCR, Dougherty et al., 2016) or the use of ethanol precipitation instead and displays both higher sensitivity and higher quantification of filtration, which has been shown to be less efficient (Spens et al., accuracy (Dingle et al., 2013; Rackiˇ et al., 2014; Zhao et al., 2016; Hinlo et al., 2017; Troth et al., 2020). For several crayfish 2016), and several studies have highlighted the benefits of ddPCR eDNA studies, consistent comparisons to CPUE data are also for eDNA detection (e.g., Doi et al., 2015a; Hunter et al., 2017; deficient or missing (Agersnap et al., 2017; Harper et al., 2018; Hamaguchi et al., 2018; Mauvisseau et al., 2019a; Brys et al., Mauvisseau et al., 2018; Rusch et al., 2020). 2020). However, while ddPCR offers absolute quantification, the Environmental DNA detection frequency compared to CPUE variation in estimated copy numbers increases at low numbers, data, and occupancy modeling, clearly shows a correlation therefore the LOQ of ddPCR is often within the same range as between the eDNA detection probability and crayfish population LOQ for qPCR (Dobnik et al., 2015). In our study, the estimated densities. Very low densities required high sampling effort while LOQ for qPCR was lower than for ddPCR applying a RSD of the few monitored high-density lakes had a very high detection 25% for estimated copy numbers. While we observed similar frequency and required only two samples for a 95% detection detection sensitivity for qPCR and ddPCR on DNA extracted probability. Dougherty et al.(2016) also observed that the eDNA

Frontiers in Environmental Science| www.frontiersin.org 10 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 11

Johnsen et al. eDNA Monitoring of Noble Crayfish

detection probability increased with crayfish population density. molecular techniques, and ultimately LOD/LOQ stringency, also However, they also found that a cumulative probability of 95% impact on the results (Hinlo et al., 2017; Hunter et al., 2017, 2019; varies with water clarity, where an increase in Secchi disk Strand et al., 2019; Klymus et al., 2020; Troth et al., 2020). Our values drastically lowered the detection probability of Faxonius results might also be prone to downstream analytical issues, e.g., a rusticus eDNA. Rice et al.(2018) found poor relationship recent report found that using the same filter type, but a different between Faxonius eupunctus eDNA-detection probability and DNA extraction procedure than we used increases DNA yield and crayfish density in a large river system. Instead, they showed detectability of crayfish and fish eDNA (Fossøy et al., 2020). a strong relationship between eDNA detection probability and Biotic factors, such as seasonality, life cycle and behavioral upstream river distance, implying that the probability of detecting traits of crayfish (Dunn et al., 2017; Stewart, 2019) are F. eupunctus eDNA increased downstream. Downstream eDNA particularly important in the context of why crayfish eDNA transport may therefore increase the detection probability more concentrations seemingly correlate poorly with crayfish densities. than the crayfish population density at the sampling site. Many biological features might impede eDNA detectability of crayfish compared to for example fish, where correlations eDNA Concentrations Compared to between population density and eDNA concentrations are seen (e.g., Doi et al., 2017; Salter et al., 2019; Stewart, 2019). Observed Relative Density of Noble A recent study report from Norway found significantly more Crayfish DNA from fish compared to crayfish in the same eDNA samples Stewart(2019) emphasizes that the factors influencing both (Fossøy et al., 2020). Crayfish commonly stay buried or hide the eDNA sources (amount of eDNA released from focal taxa) under rocks during daytime, and their hard carapace emits and eDNA sinks (removal of eDNA from the environment) little if any superfluous mucus or epidermal cells, suggesting for individual species and aquatic habitats, are manifold and reduced emission of DNA to the ambient water compared complex, both in space and time. Meaningful comparisons to many other aquatic organisms (Dougherty et al., 2016; between eDNA concentrations and species density will therefore Forsström and Vasemägi, 2016; Mauvisseau et al., 2019b). depend on in-depth knowledge of species biology and habitat, Spawning and reproduction events are known to increase eDNA combined with complex modeling (Stewart, 2019). A prerequisite signals markedly from aquatic species groups such as fish and is nevertheless a minimum of correlation in eDNA concentration amphibians (Stewart, 2019 and references therin). For crayfish, and density of the target species. We aimed to explore whether molting periods and spawning periods with elevated aggressive eDNA concentrations in the water correlate with observed behavior (Moore, 2007), have been found to elevate DNA copy relative density of noble crayfish. However, even at the highest numbers in the water column (Dunn et al., 2017; Laurendz, noble crayfish densities we had very few samples where the 2017; Harper et al., 2018). In tank experiments, Dunn et al. eDNA copy number exceeded the LOQ both for qPCR data and (2017) found a correlation between eDNA concentration and for ddPCR data. If the highly uncertain DNA copy numbers abundance of egg-bearing females. Thus, the ovigerous period in obtained below the LOQ are plotted against CPUE, a weak the late autumn could be a good period for eDNA correlations positive correlation is found. Also in other studies where a week to crayfish density. However, in cooler climate such as the positive correlation between eDNA concentrations and crayfish Nordic countries, the proportion of egg-bearing females may density estimates is observed (Dougherty et al., 2016; Cai et al., vary considerably annually, both between and within populations 2017; Larson et al., 2017), the eDNA data is rarely above what (Taugbøl et al., 1988), which may bias the results. Furthermore, would be the recommended LOQ (Klymus et al., 2020). Such if trying to compare eDNA concentrations and CPUE data data must be interpreted with great caution, and we argue that during the ovigerous period, possible correlations will probably our data is not suited for statistical correlations between eDNA be biased, as both low temperatures and trap avoidance from concentrations and relative density of noble crayfish. With our berried females will affect the catches (Abrahamsson, 1983). rather intensive sample effort in terms of sample replicates Nevertheless, compared to relatively short-term molting and and water volumes, our study is a strong documentation on reproduction events, the relatively long-lasting ovigerous period the missing correspondence between noble crayfish density and may at least be a recommendable period for eDNA sampling for noble crayfish eDNA concentrations in the water, at least for the confirming presence/absence of freshwater crayfish. In our study, time-window where we did the sampling. we monitored the crayfish after the molting season and prior Several factors contribute to why crayfish eDNA copy to the reproduction- and ovigerous season. In this respect, we number is a poor predictor of crayfish density. Abiotic and might have selected a time window that was recommendable for environmental factors (Stewart, 2019) combined with persistence achieving CPUE data that takes the crayfish biological needs into and degradation of eDNA in the lake (Dejean et al., 2011; account, but where crayfish eDNA shedding from the crayfish Barnes et al., 2014) are universal challenges in eDNA studies. population was very low. High water flow, heavy rain, clay particles and water turbidity reduce the eDNA detection likelihood (Roussel et al., 2015; Dougherty et al., 2016), which probably explains our failure to The Use of eDNA in Monitoring and detect noble crayfish eDNA after heavy rainfall in one of the Conservation of Freshwater Crayfish lakes. Sampling techniques of crayfish eDNA, including choices Environmental DNA methods are moving toward the stage of of season, depths, volume and filters, and downstream choices of being ready for, and sometimes also employed for, biodiversity

Frontiers in Environmental Science| www.frontiersin.org 11 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 12

Johnsen et al. eDNA Monitoring of Noble Crayfish

inventories and monitoring of native or invasive species eDNA concentrations and relative crayfish density with the (Leese et al., 2016; Stewart, 2019; Sepulveda et al., 2020). In chosen methods for field sampling, DNA extraction protocol Norway, comparative data obtained with eDNA monitoring and and period of sampling. Combined with previous studies, it traditional methods (cages and trapping) of native noble crayfish, seems increasingly clear that eDNA concentrations of a target introduced signal crayfish and the crayfish plague pathogen species cannot replace CPUE data or serve as a proxy for relative A. astaci, convinced the authorities to include eDNA as a crayfish density estimates. Conventional methods are still needed monitoring method (Strand et al., 2019). In 2016, eDNA was to monitor changes in population size and densities over time, officially integrated into the national crayfish plague monitoring and for providing additional information on length and weight program commissioned by the National Food Safety Authorities distribution, sex ratio, fecundity, and maturation. However, as (Vrålstad et al., 2017), and replaced a controversial cage- eDNA monitoring has proven to detect crayfish reliably even surveillance strategy with live, naïve noble crayfish for the at very low densities, this method is a powerful supplement to monitoring of crayfish plague from 2017 and onward. From monitor the presence or absence of crayfish in a larger number of 2018, the national surveillance program on noble crayfish also localities than feasible with the traditional trapping methods. If included eDNA monitoring methods, both for noble crayfish the eDNA samples are exploited for the monitoring of multiple and the invasive signal crayfish. With an integrated synergistic target species, for example alien crayfish species and diseases that approach, including joint fieldwork and sharing of results, the threaten the native crayfish species (Strand et al., 2019; Rusch two monitoring programs now focus on the pathogen (A. astaci), et al., 2020) it is also possible to implement a more holistic and and its hosts and carriers [noble crayfish and signal crayfish, see cost-efficient monitoring approach. Johnsen et al.(2019)], resulting in a more holistic monitoring approach for noble crayfish and its threats. Monitoring programs are usually put out to tender, which DATA AVAILABILITY STATEMENT to varying degrees request and emphasize competence and quality versus costs. If cost-effectiveness is regarded as important, The datasets presented in this study can be found in the compromises with the number of samples, water volumes and Supplementary Material as dataset 1–3. number of sites must often be made to deliver a competitive offer. Consequently, the monitoring programs might not be able to afford a sufficient sampling effort that ensures high AUTHOR CONTRIBUTIONS probability of detection at very low densities of crayfish, or any other rare or elusive target organism. eDNA monitoring is DS, SJ, and TV: conception and design of the study. DS, SJ, and often promoted as a cost-saving method. These savings might JR: carried out the fieldwork. DS and JR: molecular methods be at the expense of precision and the ability to provide reliable development, analyses, and results. DS and SJ: statistics and presence-absence data. In cases where the sampling effort is drafting of the manuscript. All authors contributed to content, deficient, it must be expected that the absence data covers a large editing of the manuscript, and approved the final version. portion of false negatives, and that the prospects of uncovering rare threatened species or the early invasion phase of invasive species will diminish. It is therefore of great importance to FUNDING develop smarter sampling methods, identify time windows for sampling with elevated eDNA detection probabilities, identify the This work was funded by the Norwegian Research Council molecular methods recovering the highest eDNA concentrations through the project “Targeted strategies for safeguarding from various environmental samples, and at the same time build the noble crayfish against alien and emerging threats” up acceptance and understanding among the stakeholders and (TARGET; NFR-243907), the Norwegian Veterinary Institute, authorities that quality and reliability come at a cost. Saving the Norwegian Institute for Nature Research, and the Norwegian Europe’s crustaceans, cannot be achieved by saving money on Environmental Agency. insufficient and cost-effective eDNA sampling strategies.

ACKNOWLEDGMENTS CONCLUSION We thank Bjørn Spilsberg for help with ddPCR troubleshooting, The use of eDNA in ecological research, monitoring and and Frode Fossøy and Brett K. Sandercock for help with conservation has grown tremendously in recent years. occupancy modeling. However, there are several scientists requesting a more balanced appreciation of this method’s strengths and limitations (Thomsen and Willerslev, 2015; Cristescu and Hebert, 2018; SUPPLEMENTARY MATERIAL Beng and Corlett, 2020). Among the several mentioned aspects requiring more examination is the use of eDNA concentrations The Supplementary Material for this article can be found online to predict relative density or biomass of species (Beng and at: https://www.frontiersin.org/articles/10.3389/fenvs.2020. Corlett, 2020). We found no evidence for a correlation between 612253/full#supplementary-material

Frontiers in Environmental Science| www.frontiersin.org 12 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 13

Johnsen et al. eDNA Monitoring of Noble Crayfish

REFERENCES Doi, H., Takahara, T., Minamoto, T., Matsuhashi, S., Uchii, K., and Yamanaka, H. (2015a). Droplet digital polymerase chain reaction (PCR) outperforms real- Abrahamsson, S. (1983). Trappabillity, locomotion, and diel pattern of activity of time PCR in the Detection of environmental DNA from an invasive fish species. the crayfish Astacus astacus and Pacifastacus leniusculus Dana. Freshw. Crayfish Environ. Sci. Technol. 49, 5601–5608. doi: 10.1021/acs.est.5b00253 5, 239–253. Doi, H., Uchii, K., Takahara, T., Matsuhashi, S., Yamanaka, H., and Minamoto, Ackefors, H. (1998). The culture and capture crayfish fisheries in Europe. World T. (2015b). Use of droplet digital PCR for estimation of fish abundance and Aquac. Magaz. 29, 18–67. biomass in environmental DNA surveys. PLoS One 10:e0122763. doi: 10.1371/ Agersnap, S., Larsen, W. B., Knudsen, S. W., Strand, D., Thomsen, P. F., Hesselsøe, journal.pone.0122763 M., et al. (2017). Monitoring of noble, signal and narrow-clawed crayfish using Dorazio, R. M., and Erickson, R. A. (2018). Ednaoccupancy: an r package for environmental DNA from freshwater samples. PLoS One 12:e0179261. doi: multiscale occupancy modelling of environmental DNA data. Mol. Ecol. Resour. 10.1371/journal.pone.0179261 18, 368–380. doi: 10.1111/1755-0998.12735 Barnes, M. A., Turner, C. R., Jerde, C. L., Renshaw, M. A., Chadderton, W. L., and Dougherty, M. M., Larson, E. R., Renshaw, M. A., Gantz, C. A., Egan, S. P., Lodge, D. M. (2014). Environmental conditions influence eDNA persistence in Erickson, D. M., et al. (2016). Environmental DNA (eDNA) detects the invasive aquatic systems. Environ. Sci. Technol. 48, 1819–1827. doi: 10.1021/es404734p rusty crayfish Orconectes rusticus at low abundances. J. Appl. Ecol. 53, 722–732. Beng, K. C., and Corlett, R. T. (2020). Applications of environmental DNA (eDNA) doi: 10.1111/1365-2664.12621 in ecology and conservation: opportunities, challenges and prospects. Biodivers. Dunn, N., Priestley, V., Herraiz, A., Arnold, R., and Savolainen, V. (2017). Behavior Conserv. 29, 2089–2121. doi: 10.1007/s10531-020-01980-1980 and season affect crayfish detection and density inference using environmental Berdal, K., Bøydler, C., Tengs, T., and Holst-Jensen, A. (2008). A statistical DNA. Ecol. Evol. 7, 7777–7785. doi: 10.1002/ece3.3316 approach for evaluation of PCR results to improve the practical limit of Edsman, L., Füreder, L., Gherardi, F., and Souty-Grosset, C. (2010). Astacus quantification (LOQ) of GMO analyses (SIMQUANT). Eur. Food Res. Technol. astacus. IUCN Red List Threat. Spec. 2010:e.T2191A9338388. 227, 1149–1157. doi: 10.1007/s00217-008-0830-831 Forsström, T., and Vasemägi, A. (2016). Can environmental DNA (eDNA) be used Bohmann, K., Evans, A., Gilbert, M. T. P., Carvalho, G. R., Creer, S., Knapp, for detection and monitoring of introduced crab species in the Baltic Sea? Mar. M., et al. (2014). Environmental DNA for wildlife biology and biodiversity Pollut. Bull. 109, 350–355. doi: 10.1016/j.marpolbul.2016.05.054 monitoring. Trends Ecol. Evol. 29, 358–367. doi: 10.1016/j.tree.2014.04.003 Fossøy, F., Strand, D. A., Sandercock, B. K., and Johnsen, S. I. (2020). Brys, R., Halfmaerten, D., Neyrinck, S., Mauvisseau, Q., Auwerx, J., Sweet, M., et al. Environmental DNA: Comparison of Sampling Methodology and Laboratory (2020). Reliable eDNA detection and quantification of the European weather Analysis for Detection of Crayfish and Fish in Freshwater. NINA Report 1778, loach (Misgurnus fossilis). J. Fish Biol. 1–16. doi: 10.1111/jfb.14315 Norway: Norwegian Institute for Nature Research. Bylemans, J., Gleeson, D. M., Duncan, R. P., Hardy, C. M., and Furlan, E. M. Goldberg, C. S., Turner, C. R., Deiner, K., Klymus, K. E., Thomsen, P. F., Murphy, (2019). A performance evaluation of targeted eDNA and eDNA metabarcoding M. A., et al. (2016). Critical considerations for the application of environmental analyses for freshwater fishes. Environ. DNA 1, 402–414. doi: 10.1002/edn3.41 DNA methods to detect aquatic species. Methods Ecol. Evol. 7, 1299–1307. Cai, W., Ma, Z., Yang, C., Wang, L., Wang, W., Zhao, G., et al. (2017). Using eDNA doi: 10.1111/2041-210x.12595 to detect the distribution and density of invasive crayfish in the Honghe-Hani Gomes, G. B., Hutson, K. S., Domingos, J. A., Chung, C., Hayward, S., Miller, T. L., rice terrace World Heritage site. PLoS One 12:e0177724. doi: 10.1371/journal. et al. (2017). Use of environmental DNA (eDNA) and water quality data to pone.0177724 predict protozoan parasites outbreaks in fish farms. Aquaculture 479, 467–473. Capo, E., Spong, G., Norman, S., Königsson, H., Bartels, P., and Byström, P. (2019). doi: 10.1016/j.aquaculture.2017.06.021 Droplet digital PCR assays for the quantification of brown trout (Salmo trutta) Goutte, A., Molbert, N., Guérin, S., Richoux, R., and Rocher, V. (2020). and Arctic char (Salvelinus alpinus) from environmental DNA collected in Monitoring freshwater fish communities in large rivers using environmental the water of mountain lakes. PLoS One 14:226638. doi: 10.1371/journal.pone. DNA metabarcoding and a long-term electrofishing survey. J. Fish Biol. 97, 0226638 444–452. doi: 10.1111/jfb.14383 Creed, R. P. (1994). Direct and indirect effects of crayfish grazing in a stream Hamaguchi, M., Shimabukuro, H., Hori, M., Yoshida, G., Terada, T., and community. Ecology 75, 2091–2103. doi: 10.2307/1941613 Miyajima, T. (2018). Quantitative real-time polymerase chain reaction (PCR) Cristescu, M. E., and Hebert, P. D. N. (2018). Uses and misuses of environmental and droplet digital PCR duplex assays for detecting Zostera marina DNA in DNA in biodiversity science and conservation. Ann. Rev. Ecol. Evol. Syst. 49, coastal sediments. Limnol. Oceanogr. Methods 16, 253–264. doi: 10.1002/lom3. 209–230. doi: 10.1146/annurev-ecolsys-110617-162306 10242 Deiner, K., Bik, H. M., Machler, E., Seymour, M., Lacoursière-Roussel, A., Hänfling, B., Lawson Handley, L., Read, D. S., Hahn, C., Li, J., Nichols, P., et al. Altermatt, F., et al. (2017). Environmental DNA metabarcoding: transforming (2016). Environmental DNA metabarcoding of lake fish communities reflects how we survey animal and plant communities. Mol. Ecol. 26, 5872–5895. doi: long-term data from established survey methods. Mol. Ecol. 25, 3101–3119. 10.1111/mec.14350 doi: 10.1111/mec.13660 Dejean, T., Valentini, A., Duparc, A., Pellier-Cuit, S., Pompanon, F., Taberlet, P., Harper, K. J., Anucha, N. P., Turnbull, J. F., Bean, C. W., and Leaver, M. J. (2018). et al. (2011). Persistence of environmental DNA in freshwater ecosystems. PLoS Searching for a signal: environmental DNA (eDNA) for the detection of invasive One 6:e023398. doi: 10.1371/journal.pone.0023398 signal crayfish, Pacifastacus leniusculus (Dana, 1852). Manag. Biol. Invas. 9, Dejean, T., Valentini, A., Miquel, C., Taberlet, P., Bellemain, E., and Miaud, C. 137–148. doi: 10.3391/mbi.2018.9.2.07 (2012). Improved detection of an alien invasive species through environmental Hempel, C. A., Peinert, B., Beermann, A. J., Elbrecht, V., Macher, J.-N., Macher, DNA barcoding: the example of the American bullfrog Lithobates catesbeianus. T.-H., et al. (2020). Using environmental DNA to monitor the reintroduction J. Appl. Ecol. 49, 953–959. doi: 10.1111/j.1365-2664.2012.02171.x success of the Rhine sculpin (Cottus rhenanus) in a restored stream. Front. Ecol. Demeke, T., and Dobnik, D. (2018). Critical assessment of digital PCR for Evol. 8:81. doi: 10.3389/fevo.2020.00081 the detection and quantification of genetically modified organisms. Analyt. Henriksen, S., and Hilmo, O. (2015). Norsk Rødliste for Arter 2015. Artsdatabanken, Bioanalyt. Chem. 410, 4039–4050. doi: 10.1007/s00216-018-1010-1011 Norge. Available online at: https://artsdatabanken.no/Rodliste2015/ Dingle, T. C., Sedlak, R. H., Cook, L., and Jerome, K. R. (2013). Tolerance of rodliste2015/Norge/16528 (accessed January 9, 2020). droplet-digital PCR vs real-time quantitative PCR to inhibitory substances. Hinlo, R., Gleeson, D., Lintermans, M., and Furlan, E. (2017). Methods to maximise Clin. Chem. 59, 1670–1672. doi: 10.1373/clinchem.2013.211045 recovery of environmental DNA from water samples. PLoS One 12:e0179251. Dobnik, D., Spilsberg, B., Kosir, A. B., Holst-Jensen, A., and Zel, J. (2015). doi: 10.1371/journal.pone.0179251 Multiplex quantification of 12 European union authorized genetically modified Holdich, D. M., Reynolds, J. D., Souty-Grosset, C., and Sibley, P. J. (2009). A maize lines with droplet digital polymerase chain reaction. Analyt. Chem. 87, review of the ever increasing threat to European crayfish from non-indigenous 8218–8226. doi: 10.1021/acs.analchem.5b01208 crayfish species. Knowl. Manag. Aquat. Ecosyst. 394–395:11. doi: 10.1051/kmae/ Doi, H., Inui, R., Akamatsu, Y., Kanno, K., Yamanaka,H., Takahara, T., et al. (2017). 2009025 Environmental DNA analysis for estimating the abundance and biomass of Hunter, M. E., Dorazio, R. M., Butterfield, J. S. S., Meigs-Friend, G., Nico, L. G., and stream fish. Freshw. Biol. 62, 30–39. doi: 10.1111/fwb.12846 Ferrante, J. A. (2017). Detection limits of quantitative and digital PCR assays

Frontiers in Environmental Science| www.frontiersin.org 13 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 14

Johnsen et al. eDNA Monitoring of Noble Crayfish

and their influence in presence-absence surveys of environmental DNA. Mol. to conventional surveys for measuring fish species richness. Front. Ecol. Evol. Ecol. Resour. 17, 221–229. doi: 10.1111/1755-0998.12619 8:276. doi: 10.3389/fevo.2020.00276 Hunter, M. E., Ferrante, J. A., Meigs-Friend, G., and Ulmer, A. (2019). Improving McKee, A. M., Spear, S. F., and Pierson, T. W. (2015). The effect of dilution and eDNA yield and inhibitor reduction through increased water volumes and the use of a post-extraction nucleic acid purification column on the accuracy, multi-filter isolation techniques. Sci. Rep. 9:5259. doi: 10.1038/s41598-019- precision, and inhibition of environmental DNA samples. Biol. Conserv. 183, 40977-w 70–76. doi: 10.1016/j.biocon.2014.11.031 Jerde, C. L., Mahon, A. R., Chadderton, W. L., and Lodge, D. M. (2011). "Sight- Momot, W. T. (1995). Redefining the role of crayfish in aquatic ecosystems. Rev. unseen" detection of rare aquatic species using environmental DNA. Conserv. Fish. Sci. 3, 33–63. doi: 10.1080/10641269509388566 Lett. 4, 150–157. doi: 10.1111/j.1755-263X.2010.00158.x Moore, P. A. (2007). “Agonistic behavior in freshwater crayfish,” in Evolutionary Johnsen, S. I., Skurdal, J., Taugbøl, T., and Garnås, E. (2014). Effect of mesh size Ecology of Social and Sexual Systems: Crustaceans as Model Organisms, eds J. E. on baited trap catch composition for noble crayfish (Astacus Astacus). Knowl. Duffy and M. Thiel (Oxford: Oxford University Press). Manag. Aquat. Ecosyst. 413, 1–7. doi: 10.1051/kmae/2014013 Pecoraro, S., Berben, G., Burns, M., Corbisier, P., De Giacomo, M., De Loose, Johnsen, S. I., Strand, D. A., Rusch, J., and Vrålstad, T. (2019). National Surveillance M., et al. (2019). Overview and Recommendations for the Application of Digital of Noble Crayfish and the Spread of Signal Crayfish – Presentation of Surveilance PCR. Luxembourg: Publications Office of the European Union. EUR 29673 EN. Data and Population Status. NINA Report 1590, Norway: Norwegian Institute doi: 10.2760/192883 for Nature Research. Quan, P. L., Sauzade, M., and Brouzes, E. (2018). dPCR: a technology review. Johnsen, S. I., and Vrålstad, T. (2017). Noble Crayfish (Astacus astacus) – Scientific Sensors 18:1271. doi: 10.3390/s18041271 Investigation and Proposal for Coordination of the Monitoring Programs for Racki,ˇ N., Morisset, D., Gutierrez-Aguirre, I., and Ravnikar, M. (2014). One-step Crayfish and Crayfish Plague. NINA Report 1339, Norway: Norwegian Institute RT-droplet digital PCR: a breakthrough in the quantification of waterborne for Nature Research. RNA viruses. Analyt. Bioanalyt. Chem. 406, 661–667. doi: 10.1007/s00216-013- Kelly, R. P., Port, J. A., Yamahara, K. M., Martone, R. G., Lowell, N., Thomsen, P. F., 7476-y et al. (2014). Harnessing DNA to improve environmental management. Science Reynolds, J., Souty-Grosset, C., and Richardson, A. (2013). Ecological roles of 344, 1455–1456. doi: 10.1126/science.1251156 crayfish in freshwater and terrestrial habitats. Freshw. Crayfish 19, 197–218. Klymus, K. E., Merkes, C. M., Allison, M. J., Goldberg, C. S., Helbing, C. C., Hunter, Rice, C. J., Larson, E. R., and Taylor, C. A. (2018). Environmental DNA detects a M. E., et al. (2020). Reporting the limits of detection and quantification for rare large river crayfish but with little relation to local abundance. Freshw. Biol. environmental DNA assays. Environ. DNA 2, 271–282. doi: 10.1002/edn3.29 63, 443–455. doi: 10.1111/fwb.13081 Kouba, A., Petrusek, A., and Kozák, P. (2014). Continental-wide distribution of Robinson, C. V., Webster, T. M. U., Cable, J., James, J., and Consuegra, S. (2018). crayfish species in Europe: update and maps. Knowl. Manag. Aquat. Ecosyst. Simultaneous detection of invasive signal crayfish, endangered white-clawed 413:5. doi: 10.1051/kmae/2014007 crayfish and the crayfish plague pathogen using environmental DNA. Biol. Lacoursière-Roussel, A., Côté, G., Leclerc, V., and Bernatchez, L. (2016). Conserv. 222, 241–252. doi: 10.1016/j.biocon.2018.04.009 Quantifying relative fish abundance with eDNA: a promising tool for fisheries Roussel, J. M., Paillisson, J. M., Tréguier, A., and Petit, E. (2015). The downside management. J. Appl. Ecol. 53, 1148–1157. doi: 10.1111/1365-2664.12598 of eDNA as a survey tool in water bodies. J. Appl. Ecol. 52, 823–826. doi: Laramie, M. B., Pilliod, D. S., and Goldberg, C. S. (2015). Characterizing the 10.1111/1365-2664.12428 distribution of an endangered salmonid using environmental DNA analysis. Rusch, J. C., Hansen, H., Strand, D. A., Markussen, T., Hytterød, S., and Vrålstad, Biol. Conserv. 183, 29–37. doi: 10.1016/j.biocon.2014.11.025 T. (2018). Catching the fish with the worm: a case study on eDNA detection Larson, E. R., Renshaw, M. A., Gantz, C. A., Umek, J., Chandra, S., Lodge, of the monogenean parasite Gyrodactylus salaris and two of its hosts, Atlantic D. M., et al. (2017). Environmental DNA (eDNA) detects the invasive crayfishes salmon (Salmo salar) and rainbow trout (Oncorhynchus mykiss). Parasit. Vect. Orconectes rusticus and Pacifastacus leniusculus in large lakes of North America. 11:333. doi: 10.1186/s13071-018-2916-2913 Hydrobiologia 800, 173–185. doi: 10.1007/s10750-017-3210-3217 Rusch, J. C., Mojžišová, M., Strand, D. A., Svobodová, J., Vralståd, T., and Laurendz, C. (2017). Impact of Temperature, Food Availability and Lifehistory Stages Petrusek, A. (2020). Simultaneous detection of native and invasive crayfish on the eDNA Emission from Pacifastacus leniusculus and its obligate parasite and Aphanomyces astaci from environmental DNA samples in a wide range of Aphanomyces astaci. Master thesis, University of Oslo, Oslo. habitats in Central Europe. Neobiota 58, 1–32. doi: 10.3897/neobiota.58.49358 Leese, F., Altermatt, F., Bouchez, A., Ekrem, T., Hering, D., Meissner, K., et al. Salter, I., Joensen, M., Kristiansen, R., Steingrund, P., and Vestergaard, P. (2019). (2016). DNAqua-Net: developing new genetic tools for bioassessment and Environmental DNA concentrations are correlated with regional biomass of monitoring of aquatic ecosystems in Europe. Res. Ideas Outcom. 2:e11321. Atlantic cod in oceanic waters. Commun. Biol. 2:461. doi: 10.1038/s42003-019- doi: 10.3897/rio.2.e11321 06968 Lodge, D. M., Turner, C. R., Jerde, C. L., Barnes, M. A., Chadderton, L., Egan, Sepulveda, A. J., Nelson, N. M., Jerde, C. L., and Luikart, G. (2020). S. P., et al. (2012). Conservation in a cup of water: estimating biodiversity and Are environmental DNA methods ready for aquatic invasive species population abundance from environmental DNA. Mol. Ecol. 21, 2555–2558. management? Trends Ecol. Evol. 35, 668–678. doi: 10.1016/j.tree.2020. doi: 10.1111/j.1365-294X.2012.05600.x 03.011 Mason, D. H., Dysthe, J. C., Franklin, T. W., Williams, C. L., Young, M. K., Souty-Grosset, C., Holdich, D. M., Noël, P. Y., Reynolds, J. D., and Haffner, P. McKelvey, K. S., et al. (2020). Certain detection of uncertain taxa: eDNA (2006). Atlas of Crayfish in Europe. Paris: Muséum national d’Histoire naturelle. detection of a cryptic mountain sucker (Pantosteus jordani) in the Upper Spens, J., Evans, A. R., Halfmaerten, D., Knudsen, S. W., Sengupta, M. E., Missouri River, USA. Environ. DNA 1–9. doi: 10.1002/edn3.133 Mak, S. S. T., et al. (2016). Comparison of capture and storage methods for Mauvisseau, Q., Coignet, A., Delaunay, C., Pinet, F., Bouchon, D., and Souty- aqueous macrobial eDNA using an optimized extraction protocol: advantage Grosset, C. (2018). Environmental DNA as an efficient tool for detecting of enclosed filter. Methods Ecol. Evol. 8, 635–645. doi: 10.1111/2041-210X. invasive crayfishes in freshwater ponds. Hydrobiologia 805, 163–175. doi: 10. 12683 1007/s10750-017-3288-y Stewart, K. A. (2019). Understanding the effects of biotic and abiotic factors on Mauvisseau, Q., Davy-Bowker, J., Bulling, M., Brys, R., Neyrinck, S., Troth, C., et al. sources of aquatic environmental DNA. Biodivers. Conserv. 28, 983–1001. doi: (2019a). Combining ddPCR and environmental DNA to improve detection 10.1007/s10531-019-01709-1708 capabilities of a critically endangered freshwater invertebrate. Sci. Rep. 9, 1–9. Strand, D. A., Holst-Jensen, A., Viljugrein, H., Edvardsen, B., Klaveness, D., Jussila, doi: 10.1038/s41598-019-50571-50579 J., et al. (2011). Detection and quantification of the crayfish plague agent in Mauvisseau, Q., Tönges, S., Andriantsoa, R., Lyko, F., and Sweets, M. (2019b). natural waters: direct monitoring approach for aquatic environments. Dis. Early detection of an emerging invasive species: eDNA monitoring of a Aquat. Organ. 95, 9–17. doi: 10.3354/Dao02334 parthenogenetic crayfish in freshwater systems. Manag. Biol. Invas. 10, 461–472. Strand, D. A., Johnsen, S. I., Rusch, J. C., Agersnap, S., Larsen, W. B., Knudsen, doi: 10.3391/mbi.2019.10.3.04 S. W., et al. (2019). Monitoring a Norwegian freshwater crayfish tragedy – McElroy, M. E., Dressler, T. L., Titcomb, G. C., Wilson, E. A., Deiner, K., eDNA snapshots of invasion, infection and extinction. J. Appl. Ecol. 56, 1661– Dudley, T. L., et al. (2020). Calibrating environmental DNA Metabarcoding 1679. doi: 10.1111/1365-2664.13404

Frontiers in Environmental Science| www.frontiersin.org 14 December 2020| Volume 8| Article 612253 fenvs-08-612253 December 1, 2020 Time: 20:27 # 15

Johnsen et al. eDNA Monitoring of Noble Crayfish

Strand, D. A., Jussila, J., Johnsen, S. I., Viljamaa-Dirks, S., Edsman, L., Wiik- Villacorta-Rath, C., Adekunle, A. I, Edmunds, R. C., Strugnell, J. M., Schwarzkopf, Nielsen, J., et al. (2014). Detection of crayfish plague spores in large freshwater L., and Burrows, D. (2020). Can environmental DNA be used to detect first systems. J. Appl. Ecol. 51, 544–553. doi: 10.1111/1365-2664.12218 arrivals of the cane toad, Rhinella marina, into novel locations? Environ. DNA Strand, D. A., Rusch, J., Johnsen, S. I., Tarpai, A., and Vrålstad, T. (2020). 2, 635–646. doi: 10.1002/edn3.114 The Surveillance Programme for Aphanomyces astaci in Norway 2019. Annual Vrålstad, T., Knutsen, A. K., Tengs, T., and Holst-Jensen, A. (2009). A quantitative report, Oslo: Norwegian Veterinary Institute. TaqMan (R) MGB real-time polymerase chain reaction based assay for detection Sylvestre, S., Sekela, M., Tuominen, T., and Moyle, G. (2002). Water of the causative agent of crayfish plague Aphanomyces astaci. Vet. Microbiol. quality assessment of agricultural and residential run-off using the crayfish 137, 146–155. doi: 10.1016/j.vetmic.2008.12.022 Pacifastacus leniusculus in British Columbia, Canada. Freshw. Crayfish 13, Vrålstad, T., Strand, D. A., Rusch, J. C., Toverud, Ø, Johnsen, S. I., Tarpai, A., et al. 383–395. (2017). The Surveillance Programme for Aphanomyces astaci in Norway 2016. Takahara, T., Minamoto, T., Yamanaka, H., Doi, H., and Kawabata, Z. (2012). Annual report, Oslo: Norwegian Veterinary Institute. Estimation of fish biomass using environmental DNA. PLoS One 7:e035868. Westman, K., and Pursiainen, M. (1982). Size and structure of crayfish (Astacus doi: 10.1371/journal.pone.0035868 astacus) populations on different habitats in Finland. Hydrobiologia 86, 67–72. Taugbøl, T., Skurdal, J., and Fjeld, E. (1988). Maturity and fecundity of Astacus doi: 10.1007/BF00005789 astacus females in Norway. Freshw. Crayfish 7, 107–114. Wood, S. A., Pochon, X., Laroche, O., von Ammon, U., Adamson, J., and Thomsen, P. F., Kielgast, J., Iversen, L. L., Wiuf, C., Rasmussen, M., Gilbert, Zaiko, A. (2019). A comparison of droplet digital polymerase chain reaction M. T. P., et al. (2012). Monitoring endangered freshwater biodiversity using (PCR), quantitative PCR and metabarcoding for species-specific detection in environmental DNA. Mol. Ecol. 21, 2565–2573. doi: 10.1111/j.1365-294X.2011. environmental DNA. Mol. Ecol. Resour. 19, 1407–1419. doi: 10.1111/1755-0998. 05418.x 13055 Thomsen, P. F., and Willerslev, E. (2015). Environmental DNA – An emerging tool Zhao, Y., Xia, Q. Y., Yin, Y. P., and Wang, Z. K. (2016). Comparison of in conservation for monitoring past and present biodiversity. Biol. Conserv. 183, droplet digital PCR and quantitative PCR assays for quantitative detection of 4–18. doi: 10.1016/j.biocon.2014.11.019 Xanthomonas citri Subsp citri. PLoS One 11:e0159004. doi: 10.1371/journal. Tréguier, A., Paillisson, J.-M., Dejean, T., Valentini, A., Schlaepfer, M. A., and pone.0159004 Roussel, J.-M. (2014). Environmental DNA surveillance for invertebrate species: Zimmerman, J. K. M., and Palo, R. T. (2011). Reliability of catch per unit effort advantages and technical limitations to detect invasive crayfish Procambarus (CPUE) for evaluation of reintroduction programs – A comparison of the clarkii in freshwater ponds. J. Appl. Ecol. 51, 871–879. doi: 10.1111/1365-2664. mark-recapture method with standardized trapping. Knowl. Manag. Aquat. 12262 Ecosyst. 401:07. doi: 10.1051/kmae/2011016 Troth, C. R., Burian, A., Mauvisseau, Q., Bulling, M., Nightingale, J., Mauvisseau, C., et al. (2020). Development and application of eDNA-based tools for the Conflict of Interest: The authors declare that the research was conducted in the conservation of white-clawed crayfish. Sci. Total Environ. 748, 141394. doi: absence of any commercial or financial relationships that could be construed as a 10.1016/j.scitotenv.2020.141394 potential conflict of interest. Uchii, K., Doi, H., Okahashi, T., Katano, I., Yamanaka, H., Sakata, M. K., et al. (2019). Comparison of inhibition resistance among PCR reagents for detection Copyright © 2020 Johnsen, Strand, Rusch and Vrålstad. This is an open-access article and quantification of environmental DNA. Environ. DNA 1, 359–367. doi: distributed under the terms of the Creative Commons Attribution License (CC BY). 10.1002/edn3.37 The use, distribution or reproduction in other forums is permitted, provided the Usio, N., and Townsend, C. R. (2001). The significance of the crayfish Paranephrops original author(s) and the copyright owner(s) are credited and that the original zealandicus as shredders in a New Zealand headwater stream. J. Crustacean Biol. publication in this journal is cited, in accordance with accepted academic practice. No 21, 354–359. doi: 10.1163/20021975-99990135 use, distribution or reproduction is permitted which does not comply with these terms.

Frontiers in Environmental Science| www.frontiersin.org 15 December 2020| Volume 8| Article 612253